I'm starting to wonder if Dunbar's Number is why so little empathy exists on platforms like Facebook and Twitter. Social media has been trying to scale this number, but our brains don't work like that. I wonder if, when we've surpassed our own personal limit on personal connections, our brain compensates for the overload by focusing on only what we see, e.g. text, images and video, thus making us incapable of interacting as if we're talking to another human being.
Some things in life weren't meant to scale, and there are a lot of companies trying to do the literally impossible; causing far more damage to their users than they'd ever do good.
There is actually a book co-authored by Robin Dunbar [1] that examines the implications of social media etc. It was published a few years ago as the results of the decade long Lucy Project. I highly recommend the book if you're interested in the intersection of these issues.
If I were not of the opinion that social media is more damaging than beneficial to humanity, I'd like to see a social network that limits the number of people you can interact with to ~100. Want to add a new person? Then get rid of an old one.
Similarly, hard-coded limits to contact lists, e-mail address books, etc.
It'd never work, but I'm curious what the effects would be.
> I'd like to see a social network that limits the number of people you can interact with to ~100. Want to add a new person? Then get rid of an old one.
You can do this on your own if you so care - I use Twitter in pretty close to this way (I follow 158 accounts as of today, and ~10 of them are service status accounts). Anything over that I might also be interested goes on a List and I check them whenever I'm bored (and I also cap lists at ~150 or so accounts, but none of them are even close to that).
I'm of the opinion that software shouldn't build limitations, it should only be a tool for how users want to use it. (It could be a better tool by instead telling users "Hey, you're following a lot of users, things are probably pretty noisy for you huh? Maybe you want to use a list for {these similar accounts} or something?")
> If I were not of the opinion that social media is more damaging than beneficial to humanity, I'd like to see a social network that limits the number of people you can interact with to ~100. Want to add a new person? Then get rid of an old one.
Better yet, come up with a platform that naturally incentivizes users to engage in more sophisticated interactions with a smaller number of people instead of shallow interactions with large numbers of people.
I don't think this would work since users would protest and the limits would be eventually be lifted. I have friends who already purge their FB lists now and then, so people who want to do that already can.
Also, Path [1] was social network app that had these kinds of limits back in the day.
It says in that article that their limit was 500, which might explain why there wasn't as much benefit from it as one might have hoped. Dunbar's Number is around 150; 500 is way past the point where we can effectively manage relationships.
The section of the Wikipedia page probably needs updating. The very next section describes how they went from 50 people, to 150, then lifted the limits entirely.
> ...which might explain why there wasn't as much benefit from it as one might have hoped.
I think you've got this the wrong way around. The users probably didn't like being constrained to an arbitrary number. My main point was that any service that tries to impose such limits is going to feel pressure to raise/abolish them.
You can create groups of people on Facebook. I’m pretty sure you could get something very similar to that using Facebook and some active management of lists.
That (and discipline) can make you follow that rule, but it can’t make the rest of the world follow it.
That makes it quite different, as what you think are your closest friends won’t be forced to choose between you and somebody else in their size-limited friend list.
Because of that, I doubt it would be very similar.
I think this would work brilliantly. Especially amongst teens. Probably wouldn't be a net positive for society though (I imagine there'd be more social outcasts, bullying, etc - particularly in high schools).
You would have to decouple what Path did right from what it completely did wrong (overinvest in design, CEO who cared about his own status more than company, fire sale).
I think it's just a distancing effect. People are meaner to each other over chat than in person independently of how well they know each other. Traffic is another example: since you only see other drivers as their cars, you don't think of them as people and it's easier to develop intensely aggressive feelings towards them.
there's also an interesting, ralated phenomenon (which i think was posted about here on hn) where people are biased towards a counting system that is 1, 2, 3, and then many. there are apparently cultures where this is codified into the language and larger numbers are simply grouped into "many".
Dunbar's Number seems to be the crux of Paul Graham's "You Weren't Meant to Have a Boss" without citing Dunbar's number. Just an observation.
>What's so unnatural about working for a big company? The root of the problem is that humans weren't meant to work in such large groups.
>Another thing you notice when you see animals in the wild is that each species thrives in groups of a certain size. A herd of impalas might have 100 adults; baboons maybe 20; lions rarely 10. Humans also seem designed to work in groups, and what I've read about hunter-gatherers accords with research on organizations and my own experience to suggest roughly what the ideal size is: groups of 8 work well; by 20 they're getting hard to manage; and a group of 50 is really unwieldy. [1]
>Whatever the upper limit is, we are clearly not meant to work in groups of several hundred. And yet—for reasons having more to do with technology than human nature—a great many people work for companies with hundreds or thousands of employees.
Isn't this solved by the organization structure of large companies? In a company of hundreds or thousands, nobody needs to know everyone else. You really only need to / should know your team of 5-20 and a small handfull of management above/below you.
That way, it doesn't matter if you work for a company of 20 or 2000 because you only interact with a manageable subset of all employees.
This is fine, except then your teams interactions with another team might as well be interactions with a competing company. This produces some behavior that can be difficult for outsiders (customers, the CEO, etc) to understand.
So it works well as long as your teams don't interact. And if they don't interact, why do you have a 2000 person company?
When I started at my company it was a warehouse of under 200 people. As the years went by, and we added more and more staff, things seemed more chaotic. The common consensus amongst the people who worked there for many years (before my time) was that the boss was more personal and warm. Now she sits in a closed office in such a way that discourages anyone from approaching her. The organization seemed to get more chaotic and unwieldy. Because we didn’t really organize in teams it was kind of a free for all.
I started at a company just before the dot com boom. 50 people. It was like a family. Everyone was friendly and happy, everyone knew who was responsible for what, and what they were like. At the height of the boom it grew to 700, and I can only describe the psychological environment as insanity. I bumped into new strangers in the halls every day. Nobody knew what the company's mission was anymore. Everything was failing, and blame was rampant. The managers hid. Training became next to impossible and most people didn't really understand how to do their job. It was nightmarish.
> Dunbar explained it informally as "the number of people you would not feel embarrassed about joining uninvited for a drink if you happened to bump into them in a bar"
Yeah, bizarre that this number is defined by level of embarassment. Some individuals experience huge amounts of embarassment for little reason, does that mean they know less people? Ehh. I think it's just a metaphor in the end, way to tie Dunbar's number to reality.
Bit of a tangent, but I have a fuzzy recollection from my adolescence of someone (my great uncle?) saying, "As you get older, you'll find you have fewer -- and better -- friends." At the time I had my doubts about this being a good thing, but now that I'm in my 40's it's proven true. My network continues to expand, and I'm grateful to be connected to hundreds of people, but my inner circle of true friends is smaller, better and more precious to me than it's ever been.
I think the optimal quality:quantity balance is very different across one's friendly acquaintances, professional network, and inner circle.
There's also some organizational psychologists suggesting that knowing childhood friends into adulthood helps to ground individuals towards humility and away from arrogance/cockiness.
Robin Dunbar’s papers are very clear and accessible. If you find the concept of Dunbar’s number interesting, I recommend you have a look at his other papers - lots of interesting concepts!!
I always thought of Dunbar's number as being closely related to our limits on vocabulary.
Even though 150 is pretty small, you typically "know" somebody based on how they relate to the others you know about, every permutation of who knows who, who likes who, who is similar to who and so forth. That number could be expressed as a power: 150^2 = ~22K. Most people have a vocabulary size of around 10k-30K words. It's probably a coincidence, but it's a number range that pops up a lot for medium/long term memory.
In democracies there's a tension between wanting to have have elected representatives represent smaller numbers of people so as to be more responsive on one hand, and wanting them to represent more people so that you can keep the number of legislatures not too far over Dunbar's number.
While there are complexity issues with large legislatures, I think that it's actually a good thing to keep the size of the legislature well above Dunbar's Number. You may want complete stable working relationships in an executive council or cabinet, but you really want (I would argue) a significant share of relationships in a legislative body (even within a governing majority) to be characterized by arm's-length distrust.
If you're willing to have more layers, this is reasonably easy to accomplish.
Let's call Dunbar's number 100. Have every group of 100 people elect their own L1 representative. Have every group of 100 L1's elect an L2. L4 is 100 people who represent 100,000,000, larger than most countries and any of the US states, and L5 represents 10 billion, the whole world population for some time to come.
Each L representative can have meaningful relationships with all of their immediate constituents and about half the people of their own level.
> Dunbar noted that the groups fell into three categories—small, medium and large [..] with respective size ranges of 30–50, 100–200 and 500–2500 members each.
I think it would make more sense go with the smaller category, and perhaps use a number around 49. Keep it odd to avoid a tie.
At current population of 7 billion:
144 million L1 representatives.
2.9 million L2 representatives.
60 thousand L3 representatives.
1,225 L4 representatives.
24 L5 representatives
The argument for increasing the # of levels would be that the increase in effectiveness of representatives who only have to act in the interest of $dunbar number of people below them while working amoung the $dunbar number of people on the same level to influence the person on the level above more than cancels out the decrease in effectiveness from increased layers of abstraction.
How are the groups-of-100 at each level organized? Is this plan vulnerable to gerrymandering?
My (possibly naive) fix is to allow people at each level to join any group-of-100, as long as the size remains at 100. (For one to join, one has to leave.)
At the extreme opposite of this, I started writing down all the people I knew something about, and I realized I knew something about at least 12,000 people, and probably many more than that:
And you also know some facts about at least 12,000 people, and probably many more.
Just think of your favorite music bands. You can probably list 100 bands easily. And if they average 3 people per band, that's 300 people that you know something of.
Knowing the upper limit on how many people we know is crucial for understanding the complexity of human society.
This is not about people, but knowing about the connections between people. I have no idea if say Bill Gates likes or dislikes David Letterman.
If I am going to go and say kill a Water Buffalo or tile a bathroom, then I want some people with the skills who can also work together well. That's a lot more than knowing their name and hobbies.
It's good to know the Dunbar number, but if you want to understand why human societies are so complex, you have to also know the limit. You can think of it as almost the opposite of the Dunbar number. What is the outer limit of how many people we can know something about?
Ultra weak ties have importance in practical ways. If I were hiring a high level software architect, I would expect them to have opinions about some of the big names in the history of software, maybe John McCarthy or Allan Kay or Guido von Russom or Rich Hickley. Neither of us would necessarily have any personal connections to those people, but they operate like the stars in the sky, the give us points by which we can navigate the whole entirety of the human social universe.
I have tried very hard to internalize appeal to authority as meaningless so I have mostly divorced ideas from those who had them. As such, I have never looked into big names in any field because I feel it's not just worthless but actually harmful.
I know many reasonably competent people who take the opposite approach and still get stuff done.
At no point did I invoke "appeal to authority." Rather, we navigate a social space via shared points. If we are discussing literature in the English language, it is useful to know who Shakespeare, Jane Austen and Ernest Hemmingway are. Or if we are talking about Bollywood, it is useful to know who Aishwarya Rai or Akshay Kumar are. Or we discuss basketball, it is useful to know who Lebron James is. None of that involves an appeal to authority.
Every one of those examples come from Culture not research.
If you want to talk about people that's different from getting things done. Lebron James is for example irrelevant when talking about Collage basketball so he does not impact the sport directly.
> Dunbar's number is a suggested cognitive limit to the number of people with whom one can maintain stable social relationships—relationships in which an individual knows who each person is and how each person relates to every other person.
When you're hearing "stable social relationship", you might think of the relationship to your parents, brothers, sisters, children or partner. However, the definition of "stable social relationship" also applies to your doctor, your boss, your coworkers, the lady next door who waters your flowers when you're on vacation, etc. Putting that number around 100-150 average sounds reasonable to me.
Doesn't Dunbar's explanation preclude a lot of those people?
> Dunbar explained it informally as "the number of people you would not feel embarrassed about joining uninvited for a drink if you happened to bump into them in a bar"
While the “number” itself may be similar among people, how it is utilized may differ widely.
In other words - while your brain may have the potential ability to maintain numerous connections, your life circumstances have limited the number of actual connections to only a few.
I'm very introverted, and have less than 5 close friends. This is a commonly cited feature of introversion. However, I have a rather detailed rapport, of one kind or another, with more than 100 people.
The intuition makes a lot of sense. The corporate structure and processes when everyone knows everyone (up to ~150) are very different than when they don’t. I’ve found that 150 isn’t a hard number. Companies with more remote workers tend to hit it sooner, and the ones that thrive adopt formalism earlier.
A comforting thought for many apocalyptic folks - in a postapocalyptic world, we will be operating at a scale much more in line with Dunbar's number and much better fitting our natural equipment.
Hm. Our 'natural equipment' has evolved for civilization at an amazing rate in the last 50,000 years. We're not really suited to a wild existence over most of the globe any more. Heck, take away glasses and insulin and a huge number of us die the first week.
Most people can get by without glasses in a world where they don't need to read or drive. Also most people wouldn't be diabetic on a wild forager diet.
Most of humanity are still living in rough conditions also.
Only a small percentage of humanity would die off and the rest would re-acclimate. There hasn't been all that much 'evolution' aside from some selection for intelligence.
Most of humanity is housed and fed. We shouldn't get our image of the world from movies. Nearly every country has made great strides in the human condition in the last century.
Anyway, no way the current ecosystem supports the current human population. Its only thru intense civilized effort that we've built these hives of humanity. The die-off would be immense.
And because of competition, those who are fit will dominate. So the glasses-wearing insulin-needing people will be first to go.
Philip Lieberman argues that since band societies of approximately 30–50 people are bounded by nutritional limitations to what group sizes can be fed without at least rudimentary agriculture, big human brains consuming more nutrients than ape brains, group sizes of approximately 150 cannot have been selected for in paleolithic humans.
Evolutionary psychology/nutrition is based on a lot of guesswork. We don't really know how people lived, or what timescales are important.
A lot of the early conjectures about historical developments are based on philosophical thought expirements of 19th century intellectuals. Others based on hunter-gatherers in modern times. These theories tend to be "economic" (like Malthus) and resource focused. "Agriculture created surplusses, which enables specialization..etc. underpants.. Greek temples"
In eastern Turkey there are fascinating sites from the paleo/neo-lithic transition period. Interestingly, the massive & seemingly ritual/civic site (gobleki tepeh) predates the agricultural sites. Ie, it seems a large complex pre-agricultural society existed, with many people cooperating in ways we thought were first developed by advanced agriculturalists.
One could reasonably (though not conclusively) interpret this as large-scale culture giving rise to agriculture, not the other way around.
Humans are behaviourally diverse. I imagine one would find groups of various size at all prehistorical times. For analogy, wolves are somewhat behavioural diverse. Many are solitary. Some hunt mostly rodents, and live in small packs. Some hunt mostly large game, and live in large packs. Some scavange more. There is a negotiation between wolves' inherent behavioural range and the "economics" of their lifestyle that dictates group size. Humans have a much wider range.
We don't really know much about early humans, even what they ate. We can guess though, that this varied more than it does in wolves.. at least I think so.
I recently heard a lecture on archaic sapiens in the Levant. One theorist argues they ate mostly elephants which (I imagine) required and supported large group size. We don't really know.
In any case... the reason I bring all this up is that group size, and our ability to cooperate effectively in different group sizes is just interesting. It remains (imo) central to key mystery of human origins, possibly explaining the sapiens ascention.
Yuval Noah Harari basically narrates all human history as an increasingly developing ability to cooperate in ever larger groups. This is the underlying difference between language, writing, myth, religion.. The communism-capitalism cliche is basically a debate about mechanisms of cooperation in large groups. What we are doing now (on hn) is coordinating across a large group, so we can think* together.
I think Dunbar's number (or another similar one) gives us the maximum group size for spontaneous coordination, without formalization. This is good to know.
Some things in life weren't meant to scale, and there are a lot of companies trying to do the literally impossible; causing far more damage to their users than they'd ever do good.